Semantic Management for Interoperable AI and Embedded Systems

The SMARTY Project addresses the challenge of interoperability across heterogeneous systems by leveraging semantic web technologies and structured metadata ontologies. This approach ensures seamless integration of hardware, software, datasets, and AI models in secure communication environments. 

Interoperability Framework 

Interoperability is achieved through a common metadata model that unambiguously describes digital assets—from hardware components to AI models. This model enables: 

  • Structured asset descriptions for consistent interpretation across systems. 
  • Automated data integration via machine-readable semantics (e.g., RDF, OWL). 
  • Dynamic service mediation to bridge heterogeneous implementations. 

A central asset catalog (developed in WP5) acts as a command hub, linking metadata to operational workflows (e.g., data converters, service adapters). 

Ontology Development Methodology 

The Linked Open Terms (LOT) methodology guides ontology creation through iterative cycles: 

  1. Requirements Specification: Competency questions (CQs) and domain facts are derived from workshops with stakeholders (e.g., “What firmware version is required for this hardware?”). 
  1. Implementation: Conceptual models are formalized using OWL, reusing existing ontologies (e.g., DCAT, Schema.org) to ensure compatibility. 
  1. Publication & Maintenance: Ontologies are published in multiple formats (Turtle, RDF/XML) with documentation, followed by continuous updates based on user feedback. 

Key Ontologies and Standards 

  • DCAT & Profiles: For cataloging datasets and services (e.g., ADMS for asset metadata, PROV-O for provenance tracking). 
  • Hardware-Centric Ontologies: 
  • Semantic Sensor Network (SSN): Describes sensors, system properties, and capabilities. 
  • IT Service Management Ontology (ITSMO): Links hardware/software to ownership and dependencies. 
  • Software & AI Ontologies: 
  • Software Description Ontology: Captures licensing, inputs/outputs, and variables. 
  • SPDX/CycloneDX: Standardize software bill of materials (SBOMs). 
  • AI Risk Ontology (AIRO): Aligns with EU AI Act, modeling risks, validation data, and ethical considerations. 

Applications 

This semantic framework supports: 

  • Secure edge deployments by ensuring hardware/software compatibility. 
  • AI model governance through traceable metadata (e.g., training data, performance metrics). 
  • Cross-domain interoperability in cyber-physical systems (e.g., smart mobility, defense). 

For technical details, refer to the SMARTY Project deliverables or explore the LOT methodology here. 

Share this Post